Multi-step Model Predictive Iterative Battery Peak Power Estimation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional battery management systems for electrified vehicles fail to accurately estimate peak power capability due to variations in resistor-capacitor parameters during constant current pulses, leading to errors in state of power prediction over long time intervals, which limits the vehicle's ability to utilize maximum power capacity.
Innovation Solution
The implementation of a multi-step model predictive iterative (MMPI) technique using an asymmetric equivalent circuit model with one-state hysteresis and asymmetric parameters for battery systems, which performs voltage-limited extrapolation of resistances and open-circuit voltage to determine instantaneous peak current and predicted voltage, thereby improving peak power estimation accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional SOP estimation techniques (VLOR, VLEO) are used, then the estimation process is simple and fast, but the accuracy of peak power estimation deteriorates due to variations in RC parameters during constant current pulses
Solution Approach 1:
The patent applies dynamics by making the RC parameters time-varying during the constant current pulse instead of assuming them constant. The equivalent circuit model dynamically adjusts resistance and capacitance values based on the changing state of charge (SOC) throughout the pulse duration, thereby capturing the actual battery behavior more accurately while maintaining computational feasibility through iterative solving.
Solution Approach 2:
The patent implements parameter changes by allowing the resistor and capacitor values in the equivalent circuit model to vary as functions of SOC. As the battery undergoes charging or discharging during the current pulse, the RC parameters are updated accordingly, enabling the model to adapt to the changing electrochemical state of the battery and improve peak power estimation accuracy.
2Productivity
If RC parameters are assumed constant during constant current pulse, then the calculation is computationally efficient, but the state of power prediction accuracy deteriorates over long time intervals
Solution Approach 1:
The patent applies partial action by implementing an iterative solution that performs a limited number of calculation steps to achieve sufficient accuracy without exhaustive computation. The method iteratively updates the peak current estimate and checks for convergence, stopping when the change between iterations falls below a threshold, thus balancing computational efficiency with improved accuracy over conventional single-step methods.
3Measurement precision
If voltage-limited extrapolation of resistances and open-circuit voltage is performed to determine instantaneous peak current, then the peak power estimation accuracy is improved, but the computational complexity increases
Solution Approach 1:
The patent implements feedback through an iterative calculation process where the estimated peak current is used to update the state of charge, which in turn updates the RC parameters and open-circuit voltage. The algorithm continuously refines the peak current estimate by comparing predicted terminal voltage with the voltage limit constraint, adjusting the estimate until convergence is achieved, thereby ensuring accurate instantaneous peak current prediction.
Data Source
AI summary
An electrified vehicle and method for estimating peak power of a battery system of the electrified vehicle are presented. In one exemplary implementation, the method includes receiving, at a controller of the electrified vehicle, measured current, voltage, and temperature of the battery system and determining, at the controller, operating parameters for the battery system based on the measured current, voltage, and temperature. An initial peak current at a start of a current prediction period for the battery system is determined, at the controller, based on the operating parameters, and an instantaneous peak current of the battery system is determined based on its initial peak current by performing voltage-limited extrapolation of resistances and open-circuit voltage (VLERO) of a battery model for the battery system. The battery system and an electric motor of the electrified vehicle are controlled, by the controller, based on the instantaneous peak current.